This online event, hosted by NERSC in partnership with Nvidia, will consist of a 2-day Bootcamp designed to teach users fundamental concepts and practices in deep learning, and how to effectively apply them to scientific problems. Participants will learn how to implement the primary components of deep learning workflows, including hands-on exercises training and evaluating custom deep learning models on scientific datasets.

The event will use GPU compute resources from Perlmutter, NERSC’s new cutting-edge HPC system which ranks among the top supercomputers in the world for AI.  This event has limited capacity, so please sign up early and note that active NERSC/ALCF/OLCF users will be prioritized during the registration process. 

Prerequisites

Basic experience with Python. No GPU programming or AI knowledge is required. This event has limited capacity, so please make sure that prerequisites are met before applying. 

Bootcamp Event Format

Participants are introduced to available libraries, programming models, and platforms through a mix of lectures, labs, and extensive hands-on collaboration based on real-life codes, data, and use cases. To keep the events both productive and fun, bootcamps include a competition component where teams work on mini-applications to show the best acceleration techniques.   See more information about bootcamps here. 

Agenda (tentative)

Date and Time: 9 am - 12:15 pm (Pacific time), Aug 25 - 26 (Thursday - Friday)

Day 1 (Intro): August 25, 2022 (All times are PT)

• 09:00 AM - 09:15 AM: Welcome & How to Connect to Perlmutter
• 09:15 AM - 09:45 AM: AI for Science at NERSC
• 09:45 AM - 10:45 AM: Introduction to AI (Lecture)
• 10:45 AM - 11:00 AM: Break
• 11:00 AM - 12:00 PM: CNN Primer and Keras 101 (Lecture and Lab)

Day 2 (Advanced): August 26, 2021 (All times are PT)

• 09:00 AM - 10:30 AM: Tropical Cycle Detection (Lab + Challenge)
• 10:30 AM - 10:45 AM: Break
• 10:45 AM - 12:15 PM: Steady Flow Estimation (Challenge)